本論文旨在探討閱讀檔案評量結合閱讀策略教學對於中台灣低成就國中生的影響。本研究對象為兩名八年級國中生，在七年級時因段考成績被學校歸為低成就組。這兩名學生上學期的英語成績平均相近，上課參與度高，作業確實繳交。 本研究採用Li (2008)與Yang (2007)針對國中基本學力測驗英語閱讀測驗題型提出之六大分類：確認文章主旨、引導正確結論、導出暗示推論、猜測字詞意義、尋找細部資訊、找出字詞指涉，並將此分類納為閱讀策略教學之目標。學生必須遵照閱讀檔案評量進度表，在接受閱讀策略教學同時，透過讀書心得練習策略，並透過閱讀檔案進行自評。研究為期十七週，每週一節課。為求完整了解學生閱讀與學習行為改變，並探究閱讀檔案評量結合閱讀策略教學之成效，本研究除蒐集學生閱讀檔案，亦施行仿全民英檢閱讀測驗，發放問卷，和學生進行面談，並觀察學生學習狀況。 本研究發現閱讀檔案評量結合閱讀策略教學能夠幫助低成就國中生精熟閱讀策略及提高閱讀成效。即使學生在進行檔案評量前後，對自己使用閱讀策略的認知和實際進行檔案評量的表現有所出入，學生於閱讀時進行選擇、反省、分享之情形仍較進行檔案評量前增加，充分體現檔案評量的精神。而學生閱讀行為之改變均符合Benson (2003)所談之促進學習自主原則。一旦學生學習自主性增加，學習動機隨之提高，學習反應良好，也同時負起學習責任(Cameron, 2007; Chen, 2004; Sarwar, 2002; ÜSTÜNLÜOGLU, 2009)。 本研究因此建議教育部基測委員會不應該僅以單選題型式來測驗學生閱讀能力，而應該嘗試其他題型。此外，本研究也建議英語教師設計適合學生的閱讀教材，或訂出學生閱讀分級讀本的規則，並建議編者把閱讀檔案評量與閱讀策略教學融入課本中。而按學生個別差異進行適性及彈性的調整仍有其必要性，以期發揮檔案評量結合策略教學對於學生閱讀表現之最大成效。 / The present study aims to investigate the effects of reading portfolio assessment (RPA) and reading strategy instruction on the junior high low achievers in central Taiwan. Participants in the present study were two eighth graders who were identified as low achievers according to their performance in monthly exams when in the seventh grade. They had something in common: having similar English average scores last semester, taking active participation in class, and finishing assignments by due dates. According to Li’s (2008) and Yang’s (2007) categorization of reading strategies, the Basic Competence Test (BCT) English reading comprehension tests are designed based on the following six reading strategies— identifying the main idea, drawing correct conclusions, drawing implications and inferences, determining the meaning out of the context, finding the detailed information, and finding the targets of the references. Following the RPA syllabus, the participants received reading strategy instruction, practiced the strategies through journals, and assessed themselves through reading portfolios. The observation took one class period every week for 17 weeks. With the help of simulated General English Proficiency Test (GEPT) reading comprehension tests, questionnaires, reading portfolios, conferences, and observations, the findings of the present study showed the participants’ change in reading/learning behaviors and the effectiveness of RPA/reading strategy instruction. The results indicated that RPA and reading strategy instruction may make the junior high low achievers master reading strategies and perform well in reading. Even though there is inconsistence between their understanding of their strategy use before RPA and their actual performance during RPA, they thoroughly show Katz and Johnson-Kuby’s (1996) three elements of portfolio assessment with increase in selection, reflection, and sharing when reading. The changed reading behaviors correspond to Benson’s (2003) principles for fostering learner autonomy; because of increased learner autonomy, they raise their learning motivation, have positive learning response, and take responsibility for their learning (Cameron, 2007; Chen, 2004; Sarwar, 2002; ÜSTÜNLÜOGLU, 2009). At last, the present study suggested that the BCT committee consider trying more diverse ways to test reading ability, rather than focusing only on single choice questions. Furthermore, this study suggested that English teachers design reading materials suitable for students or set up a system to get students to read higher level books, and that textbook editors include RPA and reading strategy instruction. Meanwhile, editors and teachers have to keep in mind that personal and diverse adjustments are permitted in order to have the best effectiveness on students’ learning of reading, as long as the adjustments follow guidelines for RPA.

為達成「精熟學習」的教與學過程中強調的個別化教學、學生學習的回饋、校正學生學習錯誤等高效能的教學理想。本研究自行發展一套數位化診斷系統，在不改變學校政策、班級作息、教室管理實施情況下，對準備參加科大入學測驗高工三年級的學生，在考前六週進行數學科總複習的「精熟學習配合數位化診斷」實驗教學。經由學生的作答反應，本診斷系統完成以下功能： (1)能快速的篩選出優良試題，並提供試題的難度、鑑別度、答對率與評鑑試題的品質。 (2)能診斷出個別學生與試題的差異性，並完成試題卷的信度、效度評鑑與S-P分析表。 (3)能提供完整的學習成就遷移記錄，並可歸類出學生學習潛能與整體學習表現。 (4)能彙整出學生學習成效與具體教學建議。 教師藉診斷系統所提供的資訊，對學生進行個別化校正或充實教學，期以達成精熟學習的教學理想。並在實驗教學後分析學生學習回饋單，大多數學生對實驗教學法持正面態度，且自認學習態度更積極，並建議若能提早實施精熟學習配合數位化診斷教學，應會有更好成績表現。 本研究並探討「實驗教學」的實驗A組學生，分別對「不同科別」的實驗B組學生與對「傳統教學」的控制C組學生的數學學習態度、學習成就、學習成就絕對均差的差異性，經單因子共變數分析，排除前測影響後，具體結論如下： (1)實驗A、B組與控制C組學生在數學學習態度後測得分並無明顯差異。 (2)實驗A組與控制C組在數學學習成就後測成績有明顯差異，且實驗A組分數明顯高於控制C組。 (3)不同科別實驗組的A組電機類與B組機械類學生，在數學學習成就後測成績有明顯差異，且電機類分數明顯高於機械類，但機械類排除未完成實驗的學生後，則不同科別的實驗A組、實驗B組，在數學學習成就後測成績無明顯差異。 (4)實驗A組與控制C組學生在數學學習成就後測成績的絕對均差有明顯差異，且實驗A組絕對均差小於控制C組。 最後為使實驗教學能更順利進行，建議教學前先營造愉快與舒適教學環境與良好的師生互動情誼再實施，這將會是實驗教學成功的關鍵 / Mastery Learning Strategy is a highly effective teaching method, emphasizing the importance of individualized tutoring, student feedback and learning error correction. In this study, a self-devised computerized diagnostic system was developed for the third-year vocational high school students who were preparing for the technological university entrance examination. An experimental teaching method using Mastery Learning Strategy with computerized diagnostic system was given to the aforementioned students as a math review lesson six weeks before the entrance examination, following the original school policy, class schedule or class management. From test responses, the diagnostic system achieved the following: (1)Quick selecting of good quality test items. It also evaluates the test items and provides the difficulty, discrimination, and correction rates of them. (2)Individualized diagnosis of each student vs. test items. It also completes the reliability, validity and S-P charts of the test. (3)Offering a comprehensive record of the learning process. It also analyzes the learning process of the students and sorts out their learning potential. (4)Analyzing students’ learning effects and coming up with teaching suggestions. In this experiment, the data gained from the diagnosis were used to provide each student with individualized instruction or error correction with the hope of attaining the goals of Mastery Learning. The analysis of the students’ feedback after the experimental teaching method showed that most students thought positively of the experiment, admitting that they developed a more aggressive learning attitude, and suggested that this computerized diagnostic system, had it been implemented earlier, would have helped them achieve even better grades. The students in this study consisted of three groups. Group A and B were both experimental groups except that group A were students from electrical engineering department while group B were students from mechanical engineering department. Group C was the control group receiving only conventional teaching method. Group A was compared with group B and group C respectively in terms of differences in math-learning attitude, math-learning achievement, and in the absolute mean deviation of learning achievement. By means of one-way analysis of covariance, with pre-test effects eliminated, the conclusions were as follows: (1)None of the groups exhibited significant grade differences on the math-learning attitude post-test. (2)Group A scored significantly higher on the math-learning achievement post-test than did Group C, the control group. (3)Group A of electrical engineering students scored significantly higher on the math-learning achievement post-test than did Group B of mechanical engineering students. However, if the grades of the Group B students who didn’t complete the experiment were unaccounted, no grade difference was noticeable on the math-learning achievement post-test between Group A and Group B. (4)Group A, the experimental group, had significantly lower absolute mean deviation on the math-learning achievement post-test than did Group C, the control group. Finally, it was suggested that creating a comfortable environment for teaching and learning as well as an advance establishment of a good teacher-student rapport are the key to a smooth-going and successful experimental teaching method.